Redesign Templates as a catalog-style browser

CHANGELOG.md

@@ -67,6 +67,7 @@ The format is based on Keep a Changelog and this project follows Semantic Versio
 - **Human-readable agent workflow guidance** — `AGENTS.md` now groups workflow rules into clearer sections, documents parallel worktree expectations including `origin/main` checks before commit/push requests and resync timing before validation or merge, directs agents to create a focused branch without pausing for confirmation, and asks agents to explicitly request commit approval with a suggested message and details.
 - **Templates agent composer** — The Templates authoring panel now gives the freeform request field more space and removes the preset suggestion chip.
 - **Run selection rail polish** — the Run page model chips, benchmark template selector, and response header now use clearer selected-state borders/backgrounds and denser mono text, with redundant server summary and model helper copy removed.
+- **Templates catalog shell** — `/templates` now opens as a browse-first catalog with category buckets, grouped template cards, dedicated preview state, and the existing AI-first authoring flow embedded in the new shell instead of auto-selecting the first template row.
 
 ### Fixed
 

README.md

@@ -86,7 +86,7 @@ Start with built-in benchmark templates, then create tests for one prompt, a dat
 Benchmark documents are persisted as JSON in a file-backed library and indexed into SQLite for runtime use. Built-in documents ship with the app, while user-created templates, datasets, runtime profiles, and plans are written to a local library directory so they can be restored if the database is rebuilt.
 
 **Benchmark template agent**
-Use the Templates page agent as the primary authoring flow to challenge underspecified benchmark ideas, draft runnable `chat_completion` benchmark templates, validate them against the benchmark schema, and review the live JSON before saving. The full structured editor remains available as an Advanced escape hatch.
+Use the Templates catalog page agent as the primary authoring flow to challenge underspecified benchmark ideas, draft runnable `chat_completion` benchmark templates, validate them against the benchmark schema, browse the library by category, and review the live JSON before saving. The full structured editor remains available as an Advanced escape hatch.
 
 **Benchmark runs**
 Run the same test against one model, many models, or the same model served by different inference servers.